2.1 A definition of traditional strategies
2.2 The strategy of investing in water supplies
2.3 The strategy of adjusting the species, age and sex composition of herds
2.4 The strategy of positioning livestock and conserving feed and water
2.5 The husbandry strategy
2.6 The strategy of managing and controlling water points
The term strategy occurs many times in this report. A 'strategy' is here defined as a pattern of behaviour followed by individuals, groups or organizations, which seems to have a consistent, harmonious, combined and overall effect greater than and distinct from the sum of the effects of the individual actions which constitute the behaviour. Plants and animals follow strategies without being conscious of the fact. Human beings normally intend, and are conscious of, the effects of the strategies they follow but this is not always the case; differences in strategy are better revealed by actual behaviour than by statements of intent.
We can define a 'traditional strategy' as one which has been practiced in tropical Africa and which does not require large inputs of money, equipment or skills from outside Africa. Traditional strategies to overcome shortages of water for livestock have taken a number of different forms. In this chapter we look at the nature of these forms in general, briefly quoting examples of where each has been practiced. In Chapter 3 we look in more detail at some societies which exemplify the main livestock production systems distinguished in Chapter 1 and show how each of these societies employs one or more traditional strategies to overcome water shortages.
One strategy, the 'investment strategy', followed when the technical opportunities for it exist, is to construct new water sources in water-deficit areas. In some areas this has been done on a considerable scale and with a high degree of skill. In the arid Haud-Ogaden region of Ethiopia, for example, thousands of dams, hafirs and cisterns have been constructed (Cossins, 1971a, pp. 3234). In one part of this region, over an area of 33000 km2, there are an estimated 41000 manmade water sources, i.e. 1.2 per km2 (Watson, 1973). Traditional open wells in Niger reach a depth of as much as 90 m (Bernus, 1981, p. 46) or even 100 m or more (Swift, 1979, p. 68). In some cases the existing major water sources appear to have been constructed with technical skills which have since been lost (Helland, 1980, p. 63; Rigby, 1969, p. 57). In a number of cases the technical skills or labour force, or both, were derived from outside the society which commissioned the work, either in the form of slaves (Lewis, 1978, p. 59) or on a contract basis (Bernus, 1981, p. 342). Although the transportation of water for livestock over long distances by lorry from water source to herd is not as common in Africa as it is in the Middle East (Bahaddy, 1981, p. 261; Cole, 1979, p. 15), it does occur, e.g. in southeast Ethiopia (Cossins, 1971a, p. 44) and Somalia (Lewis, 1961, p. 44). But this is not a 'traditional strategy' as defined above. Transport of water over comparatively short distances from water source to camp by donkey or camel for the use of calves or smallstock is a common and traditional practice (e.g. see Swift, 1979, pp. 147 and 154); and this form of investment in transport of water, by allowing the location of pastoral camps at a considerable distance from existing water points, can be an important alternative to investing in new water sources as a means of gaining access to grazing in water-deficit areas.
Another strategy, the 'composition strategy', concerns the appropriate composition of herds. This is a question of composition partly in terms of species, and partly in terms of age and sex. It is commonly agreed that camels, because of their low water requirements, are of the main species of domestic ruminants the best adapted to water-deficit areas, cattle the least well adapted, and that sheep and goats are in between the two. Some societies, e.g. the Kababish nomads in Sudan (Asad, 1970, p.17), believe goats have lower water requirements than sheep, while in other societies the reverse is the case; for example the Berti of the Sudan water their goats in the dry season once in 3 days but their sheep only once in 6 (Holy, 1974, p. 88).
Livestock owners react to relative water shortages by adjusting the species composition of their herds. In drier areas there tend to be relatively more camels, in wetter areas more cattle. Since different species differ in ways other than their adaptibility to water shortage, e.g. in the frequency, timing and duration of their lactation, and in the amount of labour required to tend them, it is common for pastoralists' holdings to consist of more than one species in order to obtain the best mix of advantages. However, the overall balance between different species will differ between drier, more water-deficit areas and wetter ones. For example in Kenya, under an annual rainfall of 200 mm, camels and cattle each account for about 20% of the total livestock species (in terms of biomass) and sheep and goats together for about 60%; under an annual rainfall of 500 mm cattle account for 65%, sheep and goats for 35% while camels essentially do not exist (Western, 1974). Table 3 shows the relative proportions of livestock species in two Somali clans in southeast Ethiopia, one of whose base territories is relatively drier than the other and relatively less well-equipped with wells - the only reliable dry-season water point.
Livestock owners not only adjust the species composition of their total livestock holdings but also the composition of individual herds; each species within the total holding may be herded separately so that its water requirements receive appropriate attention. This point will be discussed later.
Livestock owners also adjust the age and sex composition of individual herds within their total holdings in accordance with water requirements and adaptability to water shortage. For example the Jafarabe Fulbe pastoralists of Mali like to have a proportion of elderly intelligent steers in their herds to lead more excitable, less experienced animals away from a known water source at the end of the dry season towards better grazing (Lewis, 1978, pp. 54-55). The Borana of Ethiopia and Kenya divide their cattle into two groups. The 'dry' or 'fallow' (fora) group of animals is able to walk further and to drink less frequently and thus can exploit more distant pastures. The 'milking' (hawicha) group, because of the higher water requirements of its milking cows and calves, needs to drink more frequently and to walk less far to water (Dahl, 1979, p. 42). Not only do the 'dry' herds reach better, less overgrazed pastures by travelling further afield, but by doing so they leave more of the nearer pastures for the more sensitive milking herds.
Another strategy, the 'positioning and conservation strategy', involves two elements. One of these is the careful adjustment in space and time of the positions of different species and classes of livestock in relation to water supplies. This adjustment depends on the relative water requirements of each class and species, on the availability in space of forage in the quantity and quality appropriate to those livestock, on the location of sources of drinking water for livestock in the quantity and quality required, and on the means whereby water is extracted from its source and delivered to livestock. This strategy will be more clearly illustrated in some of the examples which follow.
Table 3. The relative proportions of livestock species in two Somali clans in southeast Ethiopia.
|
Clan |
Approx. density of dry- season water points (No. per km2) |
Proportion of total livestock (% of total biomass) |
|||
|
Camels |
Goats |
Sheep |
Cattle |
||
|
Habar Awal |
0.04 |
72 |
4 |
15 |
9 |
|
Abaskul |
2.57 |
27 |
9 |
33 |
31 |
Sources: Cossins (1971a) and Watson (1973a).
However, a few brief generalizations can be made. Where water is scarce the total livestock composite will be split up into as many herds of relatively homogeneous livestock as the amount of herding labour permits, so that each species of livestock can be managed in the most appropriate way (Swift, 1979, pp. 144-158; Cossins, 1971a, p. 45). Other things being equal (they seldom are!) milk stock will graze closer to water and dry stock furthest away from the water point, sheep and goats probably the intermediate area and cattle the nearest. However, there are differences between particular situations (e.g. see Smith, 1978, p. 85). Where calves must trek to water and are unable to cover too great a distance, most of them occupy the closest ring; but in other situations (e.g. Swift, 1979, pp. 144-158) sheep and goats occupy the nearest ring and cattle the intermediate area because calves which are too young to walk far to a water point can still have water transported to them in a camp further away.
The second element in this strategy is the conservation of the water and grazing at or around the most permanent and reliable water points ('fallback' or 'dry-season' water points). This is done for as long as possible into the dry season until the water or grazing at other less reliable water points is exhausted; coupled with this is the vacation of dry-season fullback points as soon as possible when rainfall reopens other points. The other elements in this and previously mentioned strategies are matters which essentially concern the individual livestock owner and those who collaborate with him in the labour of herding and watering (his 'herding community'). However the conservation of water and grazing at fullback water points in most of the grazing lands of tropical Africa where communal forms of land tenure prevail, is a matter where the benefit which an individual derives is dependent on many other people outside his own herding community pursuing the same strategy; otherwise the grazing around the fullback point will be used up by others before the peak of the dry season while the individual who is trying to pursue a conservation strategy is still exploiting the less reliable water points.
Conservation at fallback water points is a strategy which, according to reports, is widely practised in many arid and semi-arid parts of Africa, although seemingly to a much less extent in other zones. Among regions and societies practicing some kind of fullback conservation are: the Borana of Kenya (Dahl, 1979, p.48) and Ethiopia (Helland, 1980, p. 60); the Somali of southeast Ethiopia (Cossins, 1971a) and northern Kenya (Chambers, 1969, pp. 10-15); the Maasai of East Africa (Western, 1973, pp. 92-94); the Kababish Arabs of Sudan (Asad, 1970, pp. 17-30); the Berti of Sudan (Holy, 1974); the Fulani of the Mopti region and Niger delta (although this is a much more complex system due to both the rise and fall of flood water and the risk of damage to cultivation) (Gallais, 1975; Lewis, 1978); the Tuareg of Mali in the central Gourma region (Bourgeot, 1981), in the southeast (Smith, 1978, pp. 83-88), and in the Adrar region in the northeast (Swift, 1979, pp. 56-60); the Tuareg and Fulani in the northern Sahelian zone of Niger (Bernus, 1971, and 1979, p. 51); and to a limited extent livestock owners in southeast and eastern Botswana (Gulbransen, 1980, pp. 192-198; Fortmann and Roe, 1981, pp. 7-15).
This list of examples could, no doubt, be greatly extended by a more complete search of the literature. What is noteworthy is the small proportion of the examples in which this conservation strategy is a formal policy of a community with community-imposed rewards and sanctions for compliance. The strategy is seldom imposed by a society's rules, formally agreed on by a community or decreed by a legitimate authority; usually it is discernable mainly in the way in which people actually behave, although that behaviour may be influenced by cultural norms expressed in acceptance or disapproval by public opinion. In part this observation may be caused by the nature of the evidence on which we rely. In most cases this evidence consists of the reports of anthropologists, many of whom may not have been particularly interested in this aspect and may simply have failed to note the mechanisms by which the community enforces its policy. Nevertheless in a number of cases social scientists have looked for rules and found none.
In the Niger Delta (Gallais, 1975) and the central Gourma of Mali (Bourgeot, 1981) rules exist and are focused on the conservation of grazing. Among the Berti of Sudan (Holy, 1974, p. 107) no one may use the dry-season well before the well-master (agid al-bir) has formally opened it for the season, but there is no evidence that this opening date is determined by the exhaustion of alternative pastures and water sources. Although the Somali of northern Kenya operate a conservation system, they explicitly deny the existence of customary control or sanctions over opening or closing of grazing. In some clans day-by-day communal discipline was exercised over the process of watering but not over the dates at which water or grazing were opened for use. The only grazing controls were ones imposed by the colonial power and these have now effectively lapsed (Chambers, 1969, pp. 11-16).
In many cases the reason why pastoralists in practice conserve the grazing and water at fall-back water points is simply that these water points, being deep open wells, require a lot of labour to extract the water, and pastoralists are reluctant to supply this until absolutely necessary. Swift (1979, p. 222) has shown how a herd of 50 camels watered from a deep well needs twice the number of people (i.e. two persons) as the same herd watered from sources where livestock have direct access. In Botswana a variety of water points are used as fullback points in the dry season; in the case of some dams a form of communal control prevents their use except when other sources of water have dried up (Fortmann and Roe, 1981, p. 139). However in many cases it is the relative reliability, cost and inconvenience of using some kinds of water points which leads to their being used only at the height of the dry season when no alternative source is available (Fortmann and Roe, 1981, pp. 7-19; Bailey, 1980, pp. 9-49).
A strategy of conservation of grazing around dry-season fallback points appears also to be practiced by some wildlife populations. In this case the control mechanism which moves the animals away from the fullback point as soon as rain falls elsewhere (such emigration may start within a few hours of rainfall) is not certain but appears likely to be the relatively better quality of grazing (especially in terms of protein content) away from the fullback water points (Western, 1975, p. 274; Western, 1973, pp. 50-52 and 162164).
Livestock owners engage in some other management practices - we can call them collectively the 'husbandry strategy' - in order to overcome water shortage. In some cases selection of breeding stock, especially sires, is done in terms of characteristics associated, or thought to be associated, with ability to withstand water stress. One Somali clan in southeast Ethiopia had only white cows:
'They bred their own animals and selected for this colour and type - it was a light animal and reputed to be able to withstand drought conditions and a bad Jilal (dry season) better than coloured cows.' (Cossins, 1971a, p.70).
Some herdsmen in northern Kenya also believe that animals with light coloured coats are more drought-resistant (Lewis, 1977, p. 45). Western (1982) notes that the Maasai of East Africa recognize the relationship between the environment, productivity and coat colour, and suggests that they may be actively reinforcing prevailing selection pressures, which appear to be reflected in the tendency for the incidence of light-coat colours among cattle in Maasailand to be negatively associated with higher altitude (Finch and Western, 1977).
King (1983, pp. 34-37) has reviewed the evidence that coat colour affects the inward flow of heat to livestock from a hot environment and, thus, the degree of heat stress suffered when restriction on drinking water leads to dehydration, and has shown that in hot dry conditions light-coloured cattle are better adapted.
Maasai pastoralists also alter the hours and length of daily grazing in accordance with temperature and distance from water (Western, 1973 and Branagan, 1962, p.8). The Borana of Isiolo district in northern Kenya at the height of a drought trekked their cattle to water at night in order to reduce losses of body water in their livestock (Dahl, 1980, p. 62). Elsewhere in northern Kenya at a relatively cool time of year cattle are let out by their herdsman to graze very early in the morning and in this way are able to get all the water they need from the dew formed by the condensation of mist; as a consequence the cattle do not need to drink for up to 60 days and thus are able to graze an area that has no water point (Lewis, 1977, p. 41).
Another management variable is the duration and number of drinking episodes that take place at each visit to a watering point. When the frequency of bringing animals to the water point is restricted to once in every 3 days or more, pastoralists will often organise the routine of watering so as to allow livestock an opportunity to drink more than once at each visit (Field, 1977; Bernus, 1981, p. 30; Marty, 1972, pp. 27-29; Cossins, 1971b, p. 486; Torry, 1977, p. 10), although congestion at wells and labour shortage may prevent this. King (1983, p. 18) cites evidence that livestock that have been severely dehydrated may need more than one drink to replace lost water completely.
6 In this case watering takes place every other day.
Even where livestock owners and herders practice all the strategies discussed so far, there may still be a deficit between the amount of water the livestock in an area need and its supply. The final strategy we discuss is one for managing and controlling water points. By management is meant the organization of watering activities and maintenance in such a way that the minimum of time and water is wasted - through slow rates of extraction due to insufficient labour or other forms of energy to draw water, through quarrels and fighting about turns for watering, through fouling of water by animals or through losses from water sources or troughs. By 'control' is meant the regulation of access to a water source, and restricting this access to the number of people or livestock for which the water and surrounding grazing is adequate.
The rules and systems for managing and controlling water points differ from society to society and, within the same society, between different kinds of water sources, different seasons of the year, and sometimes between the same season in different years. The degree of management and control tends to vary with the scarcity of water, with the difficulty of extracting it, or with the amount of surrounding grass. Where neither water nor grass is scarce management and control are often perfunctory, becoming more strict as the dry season advances (Fortmann and Roe, 1981, pp. 142 and 145). In arid areas where communal systems of land and grazing tenure apply, it is usual for water in ephemeral natural pans to be unmanaged and uncontrolled; anyone within the society that has grazing rights in the area is at liberty to water his livestock at these pans (Asad, 1970, p. 21; Helland, 1980, p. 61). The water in the pans is likely to dry up more quickly, through evaporation and seepage, than animals drinking there can exhaust the water or surrounding grazing, so that the water is not a conservable resource to be kept from the livestock. However, access to reliable and, above all, permanent water, even when this is a 'gift of God', e.g. a lake or river, is more likely to be strictly controlled (Helland, 1980, p. 62; Cossins, 1971a, p. 34).
Where human labour or other resources have been invested in the development of a water resource, then, even if water is not scarce, some nominal control of access is likely to prevail, although in practice it is likely to be of a rather perfunctory nature (Swift, 1979, p. 70). For example, in some areas occupied by the Samburu in Kenya, water can be obtained simply by digging a water hole a few feet deep. The person who digs it can, in theory, refuse permission to any other person to use it; in practice doing this would create considerable bad feeling, except when water and grazing are in short supply (Spencer, 1965, p.5).
In high-rainfall areas there may be no formal control of access to water sources such as rivers, because water itself is not a particularly scarce resource. In practice it may be difficult to bring animals to water without trespassing on other people's pasture and crops so that de facto access may be controlled. On the other hand local laws and customs may prohibit riparian land owners from denying others access to water (personal observation in pre-revolution Ethiopia). In arid areas most societies regulate access to permanent water. In some cases the power to do this is vested in individuals through a concept of private property, and this power can be bought, sold or inherited. In some cases ownership is vested-in the society as a whole that grazes in that area, and in others ownership is vested in only one section of that society. In some cases within the same area and society different rules apply to different kinds of water supply. Somali pastoralists in southeast Ethiopia present an example of this (Cossins, 1971a) as do pastoralists of the same ethnic group in neighbouring northern Somalia (Mirreh, 1983), although the details of the rules differ significantly between the two areas and the account which follows applies only to southeast Ethiopia.
Particular permanent dry-season wells are usually owned by primary lineage groups (sub-clans), although several primary lineages may each have their own well or wells within a single well-field. These primary lineages will usually allow free access to members of the same clan, or occasionally the same clan family, without payment or grant of specific reciprocal rights. Members of other clans will normally have to pay in cash or kind or by the grant of reciprocal rights on a contract basis. Small hafir dams (harrs) are owned by individuals or by close family, and water may be sold from them. Large hafir dams were, in the past, owned by primary lineage groups, but there was a trend prior to 1975 to individual ownership or to ownership by syndicates which cut across primary lineage lines. Water is increasingly being sold to those who are not close relatives of the owners. Concrete-lined water cisterns (birkas or berkads) are also owned by individuals and water may be sold from them even to close relatives.
In this example, control of access to water in order to match supply with demand is regulated by a number of administrative devices. At permanent wells, if supply in a particular season falls short of demand, members of other clan families, then of the same clan family, then of other primary lineages within the clan can successively be refused access. At water sources (harrs or berkads) owned by individuals or small groups the price of water to those who are not the owners of a source can be progressively raised until they are discouraged from coming. In the early 1970s in a bad year it cost the price of a whole sheep to water 170 sheep once7.
7 On the basis of: a 30 kg sheep valued at Eth. Birr 30, water in a bad dry season sold at Eth. Birr 5 per 200 litres, sheep (50% of whom are lactating) watered every 3 days drinking the equivalent of 2.3 litres per day, i.e. 7 litres at a single watering (from King, 1983, p. 45). Cossins' assertion (Cossins, 1971a p. 69) that sheep are only given 4 litres of water every 9 days in the dry season, i.e. 0.44 litres per day, seems implausible.
Control of access to water usually only distinguishes persons with stronger or weaker claims to use a particular water point. It seldom, if ever, imposes a formal limit, by regulation, on how many stock each person may water in times of scarcity. However, other kinds of constraint may impose less formal limits. Foremost among these is the increased requirement in times of water scarcity for human labour to extract water from the source and to deliver it to livestock. At the height of a bad dry season the appetites of livestock may be relatively low and the loss of water through faeces correspondingly reduced; but this will be counterbalanced by an increased need for water for evaporative cooling necessitated both by high ambient temperatures and long treks to water (King, 1983, pp. 57-61). Livestock needs for water will, therefore, be high at precisely the time when there will be most difficulty in extracting water from wells. This difficulty arises both from the low static water level in some reliable dry-season wells (as much as 90 m below ground level in some places) and from the low yields of wells during a drought, which means that labour and livestock have to wait around for the well to recharge itself.
At some wells each herd has to produce its own labour to water its own stock. In such cases owners of herds with high stock: family labour ratios will have to make arrangements to borrow, contract, or hire labour in order to water over-large herds at peak times; alternatively they may arrange to entrust or lend their animals to be managed by people whose herds are smaller. Such arrangements can be expensive particularly in the more commercialized pastoral systems. Entrusting or lending livestock deprives the original owner of much of their products; hiring labour is not only expensive in terms of direct payments in cash or kind, but there is also the danger of incompetence or dishonesty on the part of the hired person (Bernus, 1981, p. 169; Dahl, 1979, p. 77). Rather than engage in such expense the owners of large herds may prefer to take their animals away from overcrowded water points into another region where water is less scarce and watering less expensive. For example in north Kordofan province in Sudan,
'Many families watering at the deep wells in Hamrat al-Shaikh have their main herds watered at the borewells in Um-Sunla about 50 miles west. Because although water from the Hamrat wells is free and from the borewell it is not, it is cheaper for exceptionally large herds to be watered at the borewell than it would be if they were watered by hired labour from the deep Hamrat wells. Watering at the borewell is of course quicker and less laborious than watering at the deep Hamrat wells8. But there is a risk involved: a mechanical breakdown of the borewell pump may spell disaster, as happened a year before my arrival in the geld.' (Asad, 1970, pp. 24-25).8 But a Tuareg livestock owner interviewed in Niger in 1972 denied that watering camels or sheep at boreholes involved any fewer herdsmen than watering at other sources such as open wells (Marty, 1972), even though the actual raising of the water to the ground surface is by motorised pump in the case of a borehole and by hand power in the case of an open well.
At other wells watering livestock requires the cooperation of several relatively independent herding units. An extreme example occurs among the Borana in southern Ethiopia where some open wells require a chain (team) of up to 23 strong persons to draw water. In such circumstances the owner of a herd who is unable to provide a labour force proportionate to the size of his herd will find it extremely difficult to find other herding units with whom to form a watering unit. He would, therefore, have to distribute his surplus to friends and allies or recruit extra labour through adoption, foster parenthood or through herding contracts. He will have to spend much time, and in the end many resources, in recruiting support in the well-council (cora ela) to prevent his exclusion from access to the well and a place in the watering roster (schedule of users). In extreme cases the herder of too large a herd will even bribe the well-master (abba hirega). All these arrangements have their costs, and if the herd owner fails to meet them he will be excluded from the well and his livestock will die unless he takes them off elsewhere to less labour intensive water points (Helland, 1980, pp. 63-71).
Management of water points is designed to ensure that these are efficiently used. It includes a number of activities which require a degree of coordination and organisation of effort between different individuals or groups. Wells and, to an extent, dams and hafirs need annual maintenance to remove silt and sand, to repair structures and to replace the equipment, e.g. wooden frames or windlasses, by which water is drawn (Holy, 1974, p. 107). Either daily or at the watering of each herd, minor repairs must be made to watering troughs and dung and other refuse removed so that they do not contaminate the source (Helland, 1980, pp. 66-67). Watering and labour rosters have to be drawn up so that each herd or type of animal is allocated an appropriate frequency of watering and a place in the order of watering for that day, and so that adequate labour is there when cooperation between different herding units is required. Some further rules may be needed, for example to prevent the mixing up of herds (Lewis, 1978, p. 48) or trampling of smaller animals, or to segregate sick animals (Chambers, 1969, p. 15). Some authority may be required to ensure observance of the rules and to settle disputes so that they do not lead to fighting, confusion and bloodshed.
Where water is not scarce management tends to be minimal. In regions where water is relatively abundant specialist traditional institutions for water management may not exist. In regions where water is only scarce in some seasons specialist organizations may exist but only operate at the season of scarcity. In Botswana:
'Management (of small dams9) occurs but it is management under stress at that time of year when use of the dams is critical.' (Fortmann and Roe, 1981, p. 42).9 It should be noted that these are not 'traditional' dams. They were mostly constructed with outside funds and earth-moving equipment.
In Ethiopia attempts by government to introduce a 'well-master' (abba hirega) system in areas used by the Borana pastoralists at stockponds which hold water only at the less dry times of year have been only moderately successful at best (R. Sandford, personal communication), although the system functions extremely well at the traditional dry-season permanent wells. In some regions where water is extremely scarce specialist water management institutions do not seem to have evolved. In areas used by Somali pastoralists in the Horn of Africa - areas adjacent to but more arid than the Borana areas mentioned above - no specialist institutions for water management have evolved. In southeast Ethiopia informal councils of Somali elders from among those who expect to use a well may meet to work out watering schedules (Cossins, 1971a, p.39); but amongst the Somali of northeast Kenya (except for the Borana-speaking Somali) even such informal councils appear to be lacking (Chambers, 1969).
Published information on the management of traditional water points is quite limited. There is fairly detailed information available about eastern Botswana (Fortmann and Roe, 1981), the Borana of southern Ethiopia (Helland, 1980) and the Berti of western Sudan (Holy, 1974). There is a little written information about the Somali of the Horn of Africa (Cossins, 1971a; Chambers, 1969; Lewis, 1961), the Samburu of Kenya (Spencer, 1965), and the Tuareg (Marty, 1972). For other societies our information tends to consist of a few isolated sentences in documents mainly concerned with other subjects. It is difficult to know whether lack of information about water management in societies about which good information, e.g. based on sound anthropological fieldwork, exists on other subjects is because water management is not practiced. The writer of the information may not have been interested in water management, or, although interested, he or she may not have recognized it for such. What applies to water management also applies to control of access to water. It is common to find a few sentences about the rules of 'ownership' of water points, much less common to find any information of who lays down or enforces the rules, under what circumstances the rules are rigidly enforced or relaxed, and what the word 'ownership' implies. What information we do have relates to the arid end of the ecological spectrum.
